Extended Kalman Filter Based Modified Elman-Jordan Neural Network for Control and Identification of Nonlinear Systems

Gokcen Devlet Sen; Gulay Oke Gunel; Mujde Guzelkaya

doi:10.1109/ASYU50717.2020.9259812

Extended Kalman Filter Based Modified Elman-Jordan Neural Network for Control and Identification of Nonlinear Systems

Gokcen Devlet Sen, Gulay Oke Gunel, Mujde Guzelkaya

Department of Control and Automation Engineering

Istanbul Technical University

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

9 Citations (Scopus)

Abstract

In this paper, the Extended Kalman Filter (EKF) is used for online training of a recurrent neural network (RNN) model since the EKF outperforms the first order gradient-based algorithms as a second order method. The modified Elman-Jordan Neural Network model with one hidden layer is adopted as the RNN structure. Self-connections are added in context units to investigate their effects. Then, this model is utilized for identification and online control of a nonlinear single input single output (SISO) process model. The performance of the proposed structure is evaluated by simulation results. The effects of some parameters and the number of hidden units to the performance are also examined.

Original language	English
Title of host publication	Proceedings - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781728191362
DOIs	https://doi.org/10.1109/ASYU50717.2020.9259812
Publication status	Published - 15 Oct 2020
Event	2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020 - Istanbul, Turkey Duration: 15 Oct 2020 → 17 Oct 2020

Publication series

Name	Proceedings - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020

Conference

Conference	2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020
Country/Territory	Turkey
City	Istanbul
Period	15/10/20 → 17/10/20

Bibliographical note

Publisher Copyright:
© 2020 IEEE.

Keywords

Elman-Jordan networks
Extended Kalman filter
learning control
recurrent neural networks (RNNs)
system identification

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1109/ASYU50717.2020.9259812

Cite this

Sen, G. D., Gunel, G. O., & Guzelkaya, M. (2020). Extended Kalman Filter Based Modified Elman-Jordan Neural Network for Control and Identification of Nonlinear Systems. In Proceedings - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020 Article 9259812 (Proceedings - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ASYU50717.2020.9259812

Sen, Gokcen Devlet ; Gunel, Gulay Oke ; Guzelkaya, Mujde. / Extended Kalman Filter Based Modified Elman-Jordan Neural Network for Control and Identification of Nonlinear Systems. Proceedings - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020. Institute of Electrical and Electronics Engineers Inc., 2020. (Proceedings - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020).

@inproceedings{f654d7ad45004a6a83dd22adedd3270e,

title = "Extended Kalman Filter Based Modified Elman-Jordan Neural Network for Control and Identification of Nonlinear Systems",

abstract = "In this paper, the Extended Kalman Filter (EKF) is used for online training of a recurrent neural network (RNN) model since the EKF outperforms the first order gradient-based algorithms as a second order method. The modified Elman-Jordan Neural Network model with one hidden layer is adopted as the RNN structure. Self-connections are added in context units to investigate their effects. Then, this model is utilized for identification and online control of a nonlinear single input single output (SISO) process model. The performance of the proposed structure is evaluated by simulation results. The effects of some parameters and the number of hidden units to the performance are also examined.",

keywords = "Elman-Jordan networks, Extended Kalman filter, learning control, recurrent neural networks (RNNs), system identification",

author = "Sen, {Gokcen Devlet} and Gunel, {Gulay Oke} and Mujde Guzelkaya",

note = "Publisher Copyright: {\textcopyright} 2020 IEEE.; 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020 ; Conference date: 15-10-2020 Through 17-10-2020",

year = "2020",

month = oct,

day = "15",

doi = "10.1109/ASYU50717.2020.9259812",

language = "English",

series = "Proceedings - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "Proceedings - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020",

address = "United States",

}

Sen, GD, Gunel, GO & Guzelkaya, M 2020, Extended Kalman Filter Based Modified Elman-Jordan Neural Network for Control and Identification of Nonlinear Systems. in Proceedings - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020., 9259812, Proceedings - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020, Institute of Electrical and Electronics Engineers Inc., 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020, Istanbul, Turkey, 15/10/20. https://doi.org/10.1109/ASYU50717.2020.9259812

Extended Kalman Filter Based Modified Elman-Jordan Neural Network for Control and Identification of Nonlinear Systems. / Sen, Gokcen Devlet; Gunel, Gulay Oke ; Guzelkaya, Mujde.
Proceedings - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020. Institute of Electrical and Electronics Engineers Inc., 2020. 9259812 (Proceedings - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Extended Kalman Filter Based Modified Elman-Jordan Neural Network for Control and Identification of Nonlinear Systems

AU - Sen, Gokcen Devlet

AU - Gunel, Gulay Oke

AU - Guzelkaya, Mujde

PY - 2020/10/15

Y1 - 2020/10/15

N2 - In this paper, the Extended Kalman Filter (EKF) is used for online training of a recurrent neural network (RNN) model since the EKF outperforms the first order gradient-based algorithms as a second order method. The modified Elman-Jordan Neural Network model with one hidden layer is adopted as the RNN structure. Self-connections are added in context units to investigate their effects. Then, this model is utilized for identification and online control of a nonlinear single input single output (SISO) process model. The performance of the proposed structure is evaluated by simulation results. The effects of some parameters and the number of hidden units to the performance are also examined.

AB - In this paper, the Extended Kalman Filter (EKF) is used for online training of a recurrent neural network (RNN) model since the EKF outperforms the first order gradient-based algorithms as a second order method. The modified Elman-Jordan Neural Network model with one hidden layer is adopted as the RNN structure. Self-connections are added in context units to investigate their effects. Then, this model is utilized for identification and online control of a nonlinear single input single output (SISO) process model. The performance of the proposed structure is evaluated by simulation results. The effects of some parameters and the number of hidden units to the performance are also examined.

KW - Elman-Jordan networks

KW - Extended Kalman filter

KW - learning control

KW - recurrent neural networks (RNNs)

KW - system identification

UR - http://www.scopus.com/inward/record.url?scp=85097925105&partnerID=8YFLogxK

U2 - 10.1109/ASYU50717.2020.9259812

DO - 10.1109/ASYU50717.2020.9259812

M3 - Conference contribution

AN - SCOPUS:85097925105

T3 - Proceedings - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020

BT - Proceedings - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020

Y2 - 15 October 2020 through 17 October 2020

ER -

Sen GD, Gunel GO , Guzelkaya M. Extended Kalman Filter Based Modified Elman-Jordan Neural Network for Control and Identification of Nonlinear Systems. In Proceedings - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020. Institute of Electrical and Electronics Engineers Inc. 2020. 9259812. (Proceedings - 2020 Innovations in Intelligent Systems and Applications Conference, ASYU 2020). doi: 10.1109/ASYU50717.2020.9259812

Extended Kalman Filter Based Modified Elman-Jordan Neural Network for Control and Identification of Nonlinear Systems

Abstract

Publication series

Conference

Bibliographical note

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this